AI Content & Media Evaluator — Outlier
As an AI Content & Media Evaluator at Outlier, I reviewed multimodal outputs including short-form video, audio, and images for model improvement. My tasks involved annotating inconsistencies, selecting preferred completions, and collaborating in benchmark alignment. I utilized detailed rubrics to ensure data quality and provided structured feedback for model refinement. • Evaluated video, audio, and image model outputs for perceptual and rubric-driven quality. • Annotated edge cases, inconsistencies, and ambiguous completions. • Collaborated with AI research teams on benchmark creation and alignment. • Selected preferred completions using structured criteria.